What is AgentHack?

AgentHack is a hackathon focused on developing innovative solutions using UiPath agentic technology. It's an opportunity for participants to explore and create AI-powered agents that can automate complex, non-deterministic tasks within enterprise environments. 

 

What are UiPath Agents?

UiPath Agents are AI-powered, goal-oriented capabilities that can make autonomous decisions and take actions based on reasoning, context, and data. They are designed for non-deterministic, dynamic use cases and leverage large language models (LLMs) to plan, think, act, and learn over time. 

 

How do Agents differ from traditional RPA robots?

• Agents: Goal-based, act independently, handle ambiguity, suitable for ad-hoc tasks requiring high adaptability. 

• Robots: Rule-based, act predictably, suitable for deterministic actions and routine tasks requiring high reliability. 

 

What are the core components of a UiPath Agent?

An agent consists of four core components: 

• Natural language prompt 

• Context (knowledge bases) 

• Tools (activities, automations, micro automations, or other agents) 

• Escalations (human-in-the-loop mechanisms) 

 

What types of projects can I develop during AgentHack?

You can develop various projects, such as: 

• Industry-specific agents (e.g., for legal, finance, insurance, healthcare) 

• Line-of-business agents (e.g., for HR, sales, procurement, supply chain) 

• Agent testing solutions 

• Innovative applications of agentic orchestration 

 

What resources are available to me during the hackathon?

• UiPath Automation Cloud™ with access to the Agents workspace 

• UiPath Agents Documentation and guides

• UiPath Academy courses elated to AI-powered automation 

• Access to the UiPath Community Forum for support

 

How should I approach testing my agent?

• Embed evaluation processes within broader automation contexts 

• Use the Run Agent activity in automation workflows 

• Perform end-to-end workflow simulations 

• Test integration points and cross-system communication 

• Assess performance under variable load conditions 

• Evaluate failure modes and recovery mechanisms 

 

What are some best practices for developing agents?

• Focus on well-defined tasks with limited context 

• Ensure proper error handling and escalation paths 

• Leverage existing UiPath tools and activities 

• Consider security and compliance requirements

• Design for scalability and reusability 

 

Are there any limitations I should be aware of when working with Agents?

• Agents are not suitable for all tasks, especially those requiring high accuracy or involving high stakes 

• Complex data processing may require additional specialized infrastructure 

• Agents have limited ability to learn from individual interactions in real-time 

 

How will projects be evaluated?

Projects will likely be evaluated based on: 

• Innovation and creativity 

• Practical applicability in enterprise environments 

• Effective use of UiPath Agent capabilities 

• Quality of implementation and testing 

• Potential impact on business processes 

 

Where can I find more information about UiPath Agents?

You can refer to the following resources: 

• UiPath Agents Documentation 

• UiPath Academy courses 

• UiPath Community Forum 

What are the differences between an UI Agent and a General Agents built in Agent Builder?

Aspect	UI Automation (UI Agent)	Enterprise Agents (Agent Builder)
Primary Purpose	Automate precise and resilient UI-based interactions within applications or web pages	Automate and orchestrate complex, end-to-end business processes across multiple systems, apps, and workflows
Scope of Automation	Specific, focused UI interactions (e.g., data entry, form filling, data extraction)	Broader automation involving multiple steps, decisions, and coordination across various systems
Level of Autonomy	Micro-agent approach: contextually intelligent, but limited to the task at hand	Highly autonomous: capable of reasoning, decision-making, and managing complex scenarios
Interaction Channels	Primarily User Interface (UI)-centric tasks and interactions (clicking, typing, selecting, data extraction)	UI, APIs, human-in-the-loop interactions, integrations with external systems, and conversational interactions
Role of AI	AI used primarily for resilient UI interaction (dynamic selectors, semantic targeting, adaptable UI workflows)	Advanced AI-driven capabilities: conversational agents, intelligent decision-making, advanced orchestration
Task Complexity Supported	Low-to-medium complexity: repetitive, clearly-defined UI tasks	Medium-to-high complexity: processes requiring dynamic decisions, complex logic, and orchestration
Development Experience	Simple, prompt-based, low-code/no-code development directly in UiPath Studio	Low-code visual agent orchestration, configuration, and logic-driven workflows in Agent Builder
Typical Customer Scenarios	Form automation, dynamic web-page interaction, robust data extraction, and data entry scenarios	Employee onboarding/offboarding, customer support automation, invoice processing, intelligent document processing, customer-facing virtual assistants

 

How many people should I have on my team?

We recommend having no more than four people per team, to ensure efficiency and that all team members have a chance to contribute to the project. 

 

How can I collaborate with other participants?

• Use designated collaboration platforms  

• Leverage UiPath Community Forum for broader support 

• Remember to respect intellectual property rights and adhere to the hackathon's code of conduct during collaboration. 

 

Can I submit multiple projects?

Yes! However, if selected for the finals, you must choose one project in one category to present. 

How soon will my idea be validated after submission?

Your idea will be validated within 3 business days after being submitted through the form. If your idea is not accepted in the competition, you will receive an email. 

 

How soon can I start building my solution after my idea is accepted?

Once your idea is accepted, you'll gain access to UiPath Labs and can start building your solution within a maximum of 3 business days.

 

What happens next? 

Once your idea is submitted and approved, you and your team members will be invited to a UiPath Organization in the staging environment. All of you will have organization administrator rights, allowing you to perform any necessary configurations to implement the approved idea. However, please note that you are not permitted to invite any additional users to the organization. If you need to invite other team members, assign new licenses, or face any issues, please contact us by email at community.uipath.com. 

 

Will I receive feedback on my idea or project?

We generally do not offer feedback on ideas or projects, except in specific cases

 

Will there be support during the hackathon?

You'll have access to expert mentors and community support throughout the event to help guide your development and answer technical questions.

 

Can I use AI tools or third-party APIs in my agent?

Yes, you're welcome to integrate AI tools and third-party APIs. 

 

Can I use pre-existing code or projects as part of my submission?

Yes, you can incorporate pre-existing code as part of your project. 

 

Is there a code of conduct for the hackathon?

Yes, all participants are expected to follow our community guidelines and code of conduct, which promote respect, inclusion, and fair play.

 

Are there any Terms&Conditions that I must be aware of?

Yes, you can find the general T&Cs here, the AgentHack specific T&Cs here and the MEA here. 

 

Your question wasn't answered? Write to us on the dedicated Forum thread, we will make sure to reply as soon as possible.